When foods or beverages are taken in the mouth, the resulting flavor perceptions emerge from sensory signals activated by stimulation of several senses: taste (gustation), olfaction (retronasally, as air-borne particles reach olfactory receptor through the back of the mouth), somatosensation (texture, warmth, cold, pungency), and even vision (color) and hearing (sounds of mastication). Flavor perception therefore involves multisensory interactions and integration. At the same time, flavor perceptions are also strongly influenced by the context of the food stimuli and by people's knowledge and expectations, that is, by higher-level, cognitive processes. Little is known, however, about the interaction between sensory and cognitive processes in flavor perception. It is critical to distinguish interactions taking place at different levels of flavor processing. Capitalizing on state-of- the-art psychophysical (behavioral) methods in humans, the present project tests three main hypotheses about the interactions between sensory and cognitive processing in the perception of flavor mixtures: (1) The first series of experiments (Studies 1-3) tests the hypothesis that flavor-stimulus context modifies the identification of flavor mixtures and the perceived intensitie of the components of the flavors through an adaptation-like mechanism, consistent with the hypothesis that the identification of these mixtures rests largely on the relative perceived intensities of the components. (2) The second series of experiments (Studies 4-6) tests the hypothesis that linguistic context (labels given to the stimuli before presentation) affects the identification of flavor mixtures through the operation of a general-purpose cognitive mechanism for integrating information, but has no corresponding effect on perceived flavor intensity. (3) The third series of experiments (Studies 7 and 8) tests the hypothesis that although the contrastive (adaptation-like) effects of flavor-stimulus context and the assimilative effects of linguistic context, examined in Studies 1-6, can impair accurate identification of the flavor mixtures, trial-by-trial feedback can benefit subsequent performance by providing information about the statistical properties of the stimulus distributions and about the associations of labels with flavors. Feedback may allow identification performance to improve to near optimal levels, as defined by Bayesian statistics. Besides advancing our understanding of sensory-cognitive mechanisms of flavor perception, the findings of this project bear implications for food selection and food intake. The flavors of foods and beverages are important determinants of food intake, and hence important in the regulation of nutrition and body weight. Understanding the interactions of sensory and cognitive processes in the identification of flavor components of foods and beverages, and understanding the role of feedback in aiding accurate identification, could prove valuable to developing better treatments of disorders involving swallowing, eating, nutrition, and body weight.